The disclosure relates generally to cellular ceramic articles and more particularly to cellular ceramic articles having channels that are coated with a slurry material while in a green state.
Diesel particulate filters and catalyst substrates made of extruded honeycomb ceramics are key components in modern engine after treatment systems designed to meet current and future emission legislation. Cordierite is currently the dominant material of choice for substrates and is used for diesel particulate filters as well, especially for heavy-duty applications. Cordierite-based filters are also considered for gasoline particulate filters, should future emission standards require the need for such. Other material choices include stabilized aluminum titanate (AT) and silicon carbide (SiC), either re-crystallized or Si-bonded.
Such products are typically manufactured by an extrusion process followed by drying and high temperature thermal treatment processes (firing). For filters, an additional step is required to plug honeycomb channels. Cordierite and aluminum titanate honeycombs are synthetic ceramics for which the extrusion batch primarily comprises precursor materials such as alumina, silica, titania, etc. that react during the firing step to form the finished ceramic. Additional components are added to adjust rheological properties and to aid formation of pores with the desired structure. As the final material is obtained only after chemical reaction of the raw materials during the high temperature thermal treatment, prior to that treatment, the honeycomb structures and batch material are usually referred to as being in the “green” state.
In addition to possible material differences, the honeycomb structures used today for filters and substrates also differ in the number of cells per unit area, typically expressed as cells per square inch (cpsi), the web thickness, and the porosity characteristics of the wall material, namely porosity and pore size distribution. All of the extruded products commercially manufactured today have essentially uniform porosity and pore size distribution along the wall from inlet to outlet face and across the webs from one channel to the adjacent channel, with the porous characteristics determined primarily by the composition of the green batch and the subsequent thermal treatment steps. Furthermore, such products have essentially constant web thickness from inlet to outlet, as determined by the dimensions of the extrusion die. In contrast, products with varying web thickness in the radial direction to increase mechanical strength, i.e. the web thickness increases from the center to the skin, are commercially available. This variability is typically designed through different slot sizes of the extrusion die and again does not change along the main axis of the part.
For applications requiring only a substrate (where no channels are plugged as they are in diesel particulate filters), a catalytically active material is disposed on the substrate, typically via a washcoating process. In this process, the catalytically active material is applied in the form of a slurry, with the catalyst materials being dispersed and dissolved therein. Driven by a slip casting effect, the catalyst particles are deposited primarily onto the geometric surface of the substrate with some portion actually penetrating into the substrate pore structure and acting as anchors to provide good adhesion between the coating and substrate walls. To increase the degree of adhesion, a web surface with high porosity and a tailored pore size is desirable. However, to prevent excessive penetration of the coating into the wall, where the catalyst utilization would be lower due to diffusion limitations, a web surface with lower porosity and finer pores is desirable. In addition, a very low porosity substrate has an advantage in mechanical strength.
In the case of soot filtration inside a so-called wall flow filter, as used on diesel engines today, the pressure drop increases as soot becomes trapped in the filter walls. This is undesirable from both engine operation and fuel economy perspectives. To manage the overall pressure drop of the system the filter is frequently exposed (regenerated) to conditions during which the accumulated carbon-based matter is oxidized. In general, pressure drop is determined by the geometry of the honeycomb in terms of hydraulic diameter of the channels, open area for flow and web thickness and geometric or filtration area. In addition, in the presence of soot the pressure drop increases due to the amount of soot that penetrates into the microstructure (deep bed filtration) as well as the amount of soot that accumulates on the filter wall surface (cake filtration). Due to flow restrictions in the porous wall and higher specific velocities, the impact of deep bed soot (deposited inside the porous wall) on pressure drop is significantly more pronounced compared to soot deposited as cake. It has been observed that this effect is reduced when the pore size is reduced, typically below a mean pore size of ˜10 μm. A drawback to decreasing the pore size is that the wall permeability, even without soot, decreases proportionally to the square of the pore size and linearly with wall thickness. Accordingly, a thin surface layer with both small pore size and high porosity supported by a substrate with large pore size and high porosity would serve to address at least some of these concerns.
As described above, the increase in pressure drop with accumulation of soot requires frequent regeneration of the filter and removal by oxidation of the accumulated soot. Under certain conditions, referred to as uncontrolled regeneration, the heat release during this oxidation step can be significant, resulting in an increase in the temperature inside the filter. In extreme cases, this can lead to filter damage due to thermal stresses or even melting. For filter materials, a strong correlation between the volumetric heat capacity (bulk density×specific heat capacity) and the peak temperature observed during extreme soot regeneration events has been found. For high values of the volumetric heat capacity, lower temperatures are observed. As a result, for a given material with a given specific heat capacity (J/kgK) and a given maximum temperature, a higher bulk density is required for increased soot mass limit. The latter can be achieved by either using a lower porosity material or designing a filter with lower open channel volume, i.e. thicker webs. In filter applications, the highest temperatures are usually observed at the filter exit, so having a higher density at the exit would mitigate the increases in temperature. With respect to pressure drop, however, filters with higher porosity and thinner walls are desirable. Analogous to the trade off described above for substrates, the filter designs must be optimized to balance these opposing characteristics, however such designs have not been shown to be economically obtained via a continuous extrusion process.
Catalytically active materials are now being coated not only on substrates but on some filters as well. The catalytic coating of plugged particulate filters typically is found inside the porous wall structure. This is in many cases desirable from a permeability perspective and often driven by the coating process in which the slurry is forced to flow through the walls due to the alternate plugging pattern of the filter channels. A common limitation is that any separation of catalyst functionality, i.e. due to the presence of more than one type of catalytic active material, across the web or wall is technically difficult to achieve. Having an asymmetric pore structure with small pores on one side of the wall would help to sieve/slip-cast the catalyst particles of a slurry applied from this side of the wall, preventing substantial penetration into the pore structure. An additional catalyst material could be applied from the other side of the web with resulting deposition for example into the porous wall structure. With current filter products of homogenous pore size and pore structure across the web, this is challenging at best, if not impossible.
The above application examples, although not exhaustive, demonstrate the need for substrate and filter substrate bodies with webs that have different properties either along the web from inlet to outlet face or across the web from one channel to the adjacent channels. However, such designs cannot be obtained in an economically viable manner via the continuous extrusion process. Existing methods to generate structures with such variability on a web scale are based on applying a slurry, analogous to the catalyst coating process described above, to the fired substrate body. These methods, however, require additional thermal treatment steps, generally create an interface with different thermo-mechanical properties that will result in thermal stresses, and have lower permeability as the pore structures are not continuous but rather in separate layers. The latter can be addressed to some extent by using a multitude of layers with a gradient in properties but this comes at a high manufacturing cost.